Subramani Lakshminarayanan

Come January 1, 2022, I’m starting a campaign called #Inclusive21C (or Inclusive 21st Century). I’m not going to invent new activities, but would hook my social media activities. #inclusion #TheInclusiveClub

How we understand disability shapes how we treat people. In my new video, I use my own story to explore the medical and social models. Watch and share your thoughts: youtu.be/bGLCWaebHRo #Disability #SocialModel #MedicalModel #Inclusion #Accessibility

Kevin Paul, who runs The Kevin Paul Show YouTube channel (in Tamil), is someone you can characterise as a right wing intellectual. I like his intellectual heft, but the last video he posted (Ep 206), particularly his comments on disability, has prompted me to post this video. The disability community suffered at the hands of the Nazis, and we can never let anyone reintroduce archaic understandings of disability in the name of ideology. What he said has made me shudder. Watch my video and give your comments. Of course, I would encourage you to watch episode 206 of The Kevin Paul Show if you want to do that first. The Problem I have with Kevin Paul's Video youtube.com/watch?v=HBdpFPH9…

There was a time when a neighbourhood itself could function as an accessibility system. The shopkeeper knew whose glasses had broken, who needed a letter read out, who was new to the area, and which elderly resident had not appeared for a day or two. The bus conductor recognised regular passengers and called out their stops. A neighbour might walk a child across a difficult road without making it sound like a heroic intervention. None of this was perfect. There were intrusions, assumptions and the occasional helping hand that arrived before one had asked for it. But there was also familiarity. People noticed one another. Today, we have maps that speak, groceries that arrive at the door, and apps that connect us to almost anything. These are real gains, especially for disabled people and older persons. Yet many of us live amid people whose names we do not know, in buildings where a resident can disappear from view without anyone noticing. Technology can make us more independent. But independent should not have to mean invisible, isolated or left entirely to one’s own devices. The most accessible city is not merely one with ramps, tactile paving and apps. It is one where people know how to share space, offer help without taking over, and notice when another human being may need a little support. Who were the people who made your neighbourhood livable when you were growing up? #Accessibility #DisabilityInclusion #InclusiveCities #Community #IndependentLiving #AgeingWithDignity #UniversalDesign

As a totally blind person, I find the renewed argument for segregating children with disabilities into special schools deeply troubling. Germany’s AfD has argued that specialised schools should be retained and that they should once again become the norm for pupils with special educational needs. My objection is less about the political party than about what this idea represents. The world should be moving in precisely the opposite direction. Inclusive education is not about putting a disabled child into a mainstream classroom and hoping for the best. It means making mainstream education accessible, providing reasonable accommodation, specialist teachers and individualised support, and enabling disabled and non-disabled children to grow up together. That is also the direction of the UN Convention on the Rights of Persons with Disabilities. Article 24 recognises the right to inclusive education, and the UN disability-rights committee has specifically urged Germany to move from the special-education model towards quality inclusive mainstream education. I speak about this from lived experience. As a blind person, I have to live, travel, work and interact in the same society as everyone else. Today, many disabled people must also learn to live independently. Education should prepare us for that world, not create a separate world for us. There may be children with complex individual requirements who need intensive specialist support. But specialist support does not have to mean segregation as the default. The answer to inadequate inclusion is to make inclusion work better, not to retreat from it. Making segregation the norm again would not be progress. It would be turning the clock backwards.

As someone who became blind at 18, I once saw the white cane as a sign of shame. Then I realised it was a statement of confidence. Today, I use my white cane as a symbol of independence. I am sure many blind people feel the same. Thank you, Akhilesh, for this post.

The best thing about an iPhone is that, even when an app hangs, a blind VoiceOver user does not get stranded. You can kill the app and keep using the phone. Today, Gemini became more than my iPhone 13 seemed ready to handle while I was cleaning my inbox. Shutting it down took some effort, but I never had to call a sighted person for help. That is how devices should work. I do not know how far Android and TalkBack have come because I am not using them now. But independence during a failure matters as much as accessibility when everything is working. #Apple #Accessibility #iPhone13 #IndependentLiving #ScreenReaders #VoiceOver

Tamil comedians are visionaries: V.K. ராமசாமி, நாகேஷ், மனோரமா, தேங்காய் சீனிவாசன்... Rare Comedy Collection youtu.be/1lR9FSRTZjQ?si=H4no… via @YouTube

Lately, iPhone’s dictation gets everything wrong. Accuracy levels have alarmingly dropped to the extent that it is on the brink of becoming unreliable. Especially using on apps like WhatsApp or FB and X. Is that only me or if other screen reader users also face this issue? #iPhoneAccessibility #AppleAccessibility #IOS27 #AppleVIS

There is no shortage of technology for people with disabilities. Sometimes the problem is finding out that it exists. Hadley, a US-based organisation that has been supporting people with blindness and low vision in the United States and around the world for more than a century, has launched a mobile app bringing its practical guides, videos, podcasts and community resources into one place. That sounds like a fairly ordinary app announcement. But it reminded me of something we do not discuss enough when talking about accessibility. A person who suddenly begins losing vision does not automatically acquire a directory of assistive technology along with the diagnosis. How would they know about a screen reader? Who tells them that a phone can read printed text aloud? How do they discover accessible navigation, object recognition, audiobook services or the dozens of small technological tricks that can make an enormous difference to everyday independence? Those of us who have lived with disability for years gradually accumulate this knowledge. We learn from other disabled people, mailing lists, organisations, conferences, WhatsApp groups and, quite often, through accidental discovery. Someone entering this world today starts from zero. AI is making assistive technology considerably more capable, but capability alone does not produce accessibility. People have to know that the technology exists, understand what it can do and receive enough support to begin using it confidently. Perhaps one of the less glamorous problems in accessibility technology is therefore becoming one of the more important ones. Discovery. The cleverest accessibility feature in the world is of very little use to the person who never learns that it is sitting inside the phone already in their pocket. Further reading: Hadley Helps: hadleyhelps.org #Accessibility #AssistiveTechnology #Disability #VisionLoss #DigitalInclusion #Technology

Apple’s announcements this week about Apple Intelligence understandably focused on faster chips, smarter software and new AI capabilities. What caught my attention was something that was not immediately obvious. Accessibility. For years, accessibility has largely worked on one assumption. Disabled people must learn the language of technology. VoiceOver users memorise gestures. Voice Control users memorise the exact label of a button. Screen reader users learn the structure of apps and develop workarounds when something is poorly designed or inaccessible. The burden has almost always fallen on the user. Apple’s latest accessibility announcements suggest that AI may finally begin to reverse that relationship. Instead of asking the user to adapt to the computer, the computer begins adapting to the user. VoiceOver will be able to provide richer descriptions of photographs, scanned documents and other visual content. Users can ask follow-up questions about what the camera is seeing. Voice Control is becoming more conversational. Instead of remembering the exact name of a button or its position on a numbered grid, a user can simply describe what they want: “Tap the purple folder” or “Open the guide about accessibility.” That may sound like a small change. I do not think it is. For the first time, I can see the possibility of accessibility becoming conversational rather than procedural. Instead of remembering how the computer expects me to communicate, perhaps the computer can begin understanding how I naturally communicate. That is a profound shift. But there is one part of Apple’s announcement that leaves me with mixed feelings. Many of these new AI-powered accessibility features are launching only in English-speaking markets or have language restrictions. That is a familiar story. Accessibility often arrives first. Inclusive accessibility arrives much later. For millions of people in India, language is itself an accessibility issue. Someone who is more comfortable in Tamil, Kannada, Hindi, Malayalam, Bengali, Marathi or another Indian language should not have to switch to English to enjoy the full benefits of intelligent accessibility. Accessibility is not only about disability. It is also about language. If the technology understands my accessibility needs but not my language, then it still does not fully understand me. Apple deserves credit for recognising that AI can make accessibility more natural, more intuitive and less dependent on memorisation. That is an important shift. But the real accessibility revolution will arrive only when these capabilities become available to everyone, in every major language, with the same quality and on the same timeline. Technology becomes truly accessible not when it asks us to speak its language. It becomes accessible when it learns to speak ours. #Accessibility #ArtificialIntelligence #AppleIntelligence #VoiceOver #DigitalAccessibility #InclusiveDesign #Disability #AssistiveTechnology #AI

Bengaluru has welcomed me back. After six years of Covid, personal grief, unexpected turns, unfinished plans and several conversations with life that did not exactly go as scheduled, I am back in this city as a resident. Bengaluru has always had a strange way of making me feel independent, slightly lost, occasionally overcharged by auto drivers, but somehow at home. The city has changed. I have changed. Even the traffic has probably evolved into a separate civilisation. But some things remain. The air still carries memory. The Metro still gives hope. The city still has that familiar mix of chaos, kindness, ambition, filter coffee and sudden rain. This is not just a return to a place. It feels like returning to a version of myself that Bengaluru once helped me discover. So here I am, back in the city that taught me freedom, friendship, survival and the fine art of waiting patiently in traffic while pretending to be philosophical. Hello again, Bengaluru. Be gentle. Or at least give me good Wi-Fi. #Bengaluru #Bangalore #BackToBengaluru #NewBeginnings #CityOfMemories #LifeAfterGrief #HomeAgain #IndependentLiving #UrbanLife #PersonalJourney

Imagine if parents looked at their six-year-old and announced: “We have decided not to mix love with duty, exhaustion, and guilt. We have enrolled you in a full-time residential facility. We love you deeply, which is why we are leaving your upbringing to licensed professionals. We will visit every second Sunday.” The collective outrage would be deafening. We would call it abandonment masquerading as rational detachment. Yet when the roles reverse, when the ones needing feeding, bathing, and company are frail elders instead of toddlers, the conversation gets repackaged as modern enlightenment. Perhaps that is the uncomfortable truth: this reversal is not hypothetical; it is already the blueprint. For decades, the modern global economy has nudged parents into outsourcing childhood. We trade presence for provision. We hand toddlers to daycares at four months, ship adolescents off to competitive boarding academies, and hand children screens to buy silence, calling it independence. A generation raised on structured shifts and institutional care naturally matures into adults who view caregiving as an operational burden best solved by professional infrastructure. If your formative experience of affection was scheduled between work hours, why would you treat your ageing parents any differently? Emotional distance becomes an inherited reflex. The fatal flaw in adopting the Nordic lens is the illusion of universal export. Norway’s eldercare model functions because of a multi-trillion-dollar sovereign wealth fund, world-class municipal infrastructure, institutional transparency, and an egalitarian social contract where cradle-to-grave welfare is funded by immense tax rates. In that ecosystem, stepping back is not an act of neglect; the state genuinely catches the falling. Transplanting that worldview to India ignores reality. In a country where public geriatric healthcare is strained, elder-care facilities range from unaffordable luxury resorts to bleak warehousing, and regulatory oversight is brittle, filial presence is not a sentimental burden; it is the primary social security system. It is easy to intellectually deconstruct duty as “guilt” from the comfort of an Oslo apartment. But stripping away duty from love reduces human bonds to convenience. Love without obligation is a luxury reserved for the young and self-sufficient. Real care, the kind that built us when we were helpless, is rarely sterile or convenient; it is messy, exhausting, and rooted precisely in showing up. moneycontrol.com/news/trends… #ElderCare #Parenting #FilialDuty #IndianDiaspora #CultureAndCare #AgeingParents

There is something deeply strange about the new “humanise AI writing” industry. First, people are told not to worry too much about learning how to write. Just use AI. Then the AI produces text that sounds too smooth, too predictable, too evenly polished. So another tool is sold to make that text sound less artificial. That is the loop now. Generate with AI. Detect the AI flavour. Remove the obvious signs. Audit the removal. Rewrite again until the piece can pass as human. The issue is not the use of tools. Writers have always used tools. Dictation, spellcheck, style guides, editors, research databases and transcription software can all help. The issue begins when the tool replaces the learning. Good writing is not only sentence variation. It is not just rhythm, vocabulary or paragraph length. It is judgment. It is memory. It is knowing when a sentence must be short because the point is sharp. It is knowing when a paragraph must slow down because the reader needs to feel the weight of the idea. AI can imitate some of this. It can be asked to sound warmer, rougher, sharper or more personal. But imitation is not experience. If the first draft has no human thought in it, “humanising” it later does not magically create one. It only hides the absence better. The real question is not whether AI can help us write. Of course it can. The question is whether we are using it to strengthen our writing, or to avoid developing one of the most human skills we have. Because if we outsource the draft, the voice and the audit, what exactly is left of the writer? #AIWriting #Writing #DigitalLiteracy #ArtificialIntelligence #ContentCreation #HumanCreativity #MediaLiteracy

"To write or AI write?" That is the question, as Hamlet in the AI age might have asked. I ask a different question. When I give a passage for proofreading and it comes back with an odd, slightly confusing word, like quietly, I ask myself whether that word belongs to what I write. I have been writing since before the AI age, and I know for a fact that I do not reach for certain words, however fanciful they might appear. The moment I started spotting quietly in other people's writing, again and again, I decided it was time to say no to that word, much as Stephen King banned the word zestful after seeing it overused in another novelist's work. There is also the adverb rule, that you should avoid adverbs like frequently, is another thing that pushed me to strike quietly from the list. Here is an exercise, try it yourself. Write on your own, whatever it is. Then go back and check whether a word made it in because it is fashionable, or because it sounds professional, or because it sounds honourable. Words like quietly make you feel that way about yourself. When you stop reaching for those words, you are in the clear. Using AI for brainstorming or for proofreading is one thing. Using AI to write is unacceptable. So here is what the Hamlet of today should learn. "You write or AI write?" That is not the question. It is always you write and AI corrects, or you write and you fight with AI for your brownie points. #AI&Writing #AI&Content #AI #ArtificialIntelligence

Make fun of AI all you want. Be derisive. Pour scorn on it. But AI is becoming more pervasive by the day. Take LinkedIn, for example. For years, many of us wrote to be found by people or by Google. Increasingly, we may also be writing to be found — and cited — by AI assistants such as ChatGPT, Gemini and Perplexity. LinkedIn itself is now publishing guidance on making posts and articles easier for AI systems to retrieve and reference. The emphasis is not on gaming algorithms but on writing with clear structure, meaningful headlines, direct answers and genuine expertise. LinkedIn's own research suggests that AI systems increasingly favour content that is educational, well structured and authoritative, rather than simply optimised for traditional search. citeturn0search0turn0search1 For journalists, researchers and subject matter experts, this raises an interesting question. If AI becomes one of the primary ways people discover information, should we start thinking not just about SEO (Search Engine Optimisation), but also about how our work is interpreted, attributed and cited by AI? This is not about stuffing posts with keywords. It is about creating original, well structured, trustworthy content that both humans and AI can understand. The next frontier may not simply be writing for readers. It may be writing for readers as well as the AI systems that increasingly guide them to us. Useful resources: • LinkedIn: How to Maximize AI Visibility for Your LinkedIn Posts • LinkedIn: Introducing a Guide to Optimizing Your Owned Content for AI Discovery • LinkedIn: How to Optimize Your Owned Content for AI Search (Guide) • LinkedIn AI Search Resource Hub #ArtificialIntelligence #LinkedIn #Journalism #ContentStrategy #GenerativeAI #AI #Writing #DigitalPublishing

Would you be comfortable talking to someone if you knew the glasses on their face could be recording you? That question is becoming less hypothetical. The Guardian has reported a growing backlash against Meta’s smart glasses over privacy and non-consensual recording. Restaurants, pubs and theatres in the UK have begun banning them, while concerns are growing about people being recorded without their knowledge. The problem is deceptively simple: we usually know when somebody points a phone camera at us. A camera embedded in a pair of glasses is much harder to notice. Meta’s glasses have an LED that lights up while recording. But is a small indicator light enough to constitute meaningful notice — let alone consent? There is another side to this that matters greatly to me. For blind people, AI-powered smart glasses can be genuinely transformative. They can describe surroundings, read text, identify objects and provide access to information that sighted people obtain almost effortlessly. So how do we preserve that independence without creating a world in which everyone has to wonder whether the person opposite them is recording them? Where should the boundaries be — workplaces, hospitals, restaurants, classrooms or private conversations? I do not think the answer is to reject smart glasses. Nor should accessibility become an argument for dismissing legitimate privacy concerns. The real challenge is designing technology in which accessibility and meaningful consent can coexist. Because if smart glasses become mainstream, we should probably settle some of these questions before the camera on someone’s face becomes something we simply stop noticing. The Guardian: theguardian.com/technology/2… #SmartGlasses #Meta #ArtificialIntelligence #AI #Privacy #Accessibility #Blindness #AssistiveTechnology #DigitalPrivacy #InclusiveTechnology

There is another side to the debate about AI taking away jobs that deserves attention: the new jobs and positions of authority it may be creating. Echoloc Research analysed job postings across a database of more than 17 million listings and identified 1,142 companies currently recruiting for senior AI leadership positions — Head, VP, Director or Chief of AI. An important caveat: this is primarily US-focused research. It should not be read as a picture of the global labour market, much less as evidence of what is already happening in India. Even within the US, job advertisements tell us what companies say they are looking for. They do not necessarily tell us whether those positions were eventually filled, how much authority the new executives will have, or whether the organisational transformation anticipated in the job description will actually happen. So this is evidence based on recruitment signals, not a definitive measurement of how companies are restructuring themselves around AI. But the signals themselves are fascinating. According to the research, 69% of the companies recruiting AI leaders are outside the technology sector. They include banks, insurers, healthcare and pharmaceutical companies, and consumer businesses. Even more strikingly, 95% of the companies identified had apparently never advertised an AI leadership position before 2026. The language of the advertisements is revealing, too. Terms such as “enablement” and “transformation” appear more frequently than “engineering”. That suggests companies may be looking for something beyond people who can simply build AI systems. After several years of employees experimenting with ChatGPT, Copilot and other tools, organisations may be entering another stage: somebody at the senior level now has to decide where AI belongs in the business, which workflows should change, where human judgement remains essential, and whether the investment is actually producing results. There is some independent evidence pointing in the same direction. Indeed Hiring Lab has reported that more than six in ten US job postings containing AI in the job title are now outside traditional technology companies. LinkedIn has previously documented the growth of Head of AI positions as well. None of this negates legitimate concerns about AI-driven job displacement. Nor does it prove that AI is creating more jobs than it eliminates. What it does suggest is that the employment effects of AI may be more complicated than the familiar question: “Which jobs will AI replace?” We may simultaneously be watching some tasks disappear, existing occupations change and an entirely new layer of organisational responsibility emerge. And although this particular research does not establish that the same thing is happening in India, it gives us something useful to watch. Technology and management trends that become established in large US companies often travel. If this pattern continues, over the next few months and quarters it will be worth watching Indian banks, insurers, healthcare companies, manufacturers, retailers and other non-tech businesses. The interesting question may not simply be how many AI engineers they hire. Watch for the Heads of AI, AI transformation leads, AI enablement heads and similar positions. Those job advertisements could tell us something about the next phase of AI adoption before the companies themselves begin talking publicly about it. Original Echoloc research (access may be restricted/paywalled): echoloc.ai/research/whos-hir… #ArtificialIntelligence #AI #FutureOfWork #Jobs #Hiring #AIJobs #Leadership #Workplace #India #Technology #EmploymentTrends

AI detection tools are moving into publishing, literary prizes and writing platforms with more confidence than their accuracy warrants. The consequences are serious for Indian English writers whose linguistic patterns may be treated as anomalous, as well as disabled writers who use AI features for access rather than prose generation. In this article, I examine the limits of detection and describe how I use AI as a proofreader and accessibility enhancer. Drawing on my experience, I suggest that publishers should consider the writer’s drafts and editorial process instead of relying on a detector percentage. #AIDetection #AIPublishing #IndianEnglish #AssistiveTechnology #DisabilityInclusion